Presenter: Andreas Grimstvedt, SINTEF Industry
Partial least square regression (PLS-R) methodology have been applied on dataset containing FTIR spectra and densities for a large set of MEA solvent samples.
The prediction capabilities for the major compounds were tested on different set of realistic degraded solvent samples ranging from bench scale experiments to pilot plant campaigns.
Generally, the methods showed good results with exception of samples with high amount of heat stabile salts (HSS). For the real samples, sample residuals (i.e. part of data not fitted by the model) were also studied, and a clear correlation between the residuals and HSS level in samples were observed.
Online techniques based on PLS-R and FTIR should be an attractive alternative for monitoring the major solvent compounds in CO2 capture plants using amine solvents, as this will reduce cost of chemical analysis and could be an important tool for implementation of control strategies for energy saving in the process.
The webinar is open to NCCS partners, only.
